Model specification and prior predictive simulation for ONS smoking and COVID-19 study

Selecting Knots

Blah blah blah, we selected knots using the below data

plotly::plot_ly(
  x = case_data$date,
  y = case_data$new_case,
  type = "bar"
)

The knot dates are shown

ggplot(case_data, aes(x = date, y = new_case, fill = Segment_group)) +
  geom_col() +
  geom_smooth(method = "lm", colour = "black") +
  theme_minimal() +
  labs(x = element_blank(),
       y = "Number of cases",
       fill = "Segment changes")